A robust precoding for RF mismatched massive MIMO transmission

Due to radio-frequency (RF) circuit mismatch, the channel reciprocity of time-division duplex massive multiple-input multiple-output system is impaired. Under this condition, there exist several different approaches for base station (BS) to obtain the downlink (DL) channel estimate based on the minimum mean-square-error (MMSE) estimation method. We show that with the RF mismatch parameters BS will obtain the same DL channel estimates via these different approaches. As the DL channel estimate is usually imperfect, we propose an MMSE based multiuser precoding technique, which is robust to the estimation error. Furthermore, we derive an asymptotic approximation of the ergodic sum rate for the robust MMSE precoding using the large dimensional random matrix theory, which is tight as the numbers of both antennas at BS and user terminals tend to infinity with a fixed non-zero and finite ratio. Our results are verified by the simulations.